The goal of the proposed project is an in-depth analysis of the level, component structure, and expression of hostility in middle-aged and older men and women, focusing on interview-assessed hostility that has been shown to be predictive of adverse health outcomes. Hostility will be assessed with multiple methods: self-report, content analysis of speech, clinical judgments of behavior, ratings of facial expressions, and reactions to experimental stimuli. Four interrelated studies, some experiments involving manipulation of the social setting, will assess hostile behavior during social interaction and will address the following questions: 1. Are there gender differences in hostility that might explain gender differences in health? 2. Are there age differences in hostility that might explain why hostility differentially affects the health of older and middle-aged individuals? 3. Are there differences in hostility between patients with coronary artery disease and nonpatient controls matched on age and sex? If so, it would help explain the role that hostility plays in the development of coronary disease. 4. What self-report measures of hostility best correlate with interview- assessed hostile behavior and are there personality variables that mediate between self-report and behavior to effect the size of these correlations? The data designed to answer this questions will provide the basis for the development of better measurement techniques. 5. What is the stability of hostility over time?